Bismuth-doped fiber amplifiers offer an attractive solution for meeting continuously growing enormous demand on the bandwidth of modern communication systems. However, practical deployment of such amplifiers require massive development and optimization efforts with the numerical modeling being the core design tool. The numerical optimization of bismuth-doped fiber amplifiers is challenging due to a large number of unknown parameters in the conventional rate equations models. We propose here a new approach to develop a bismuth-doped fiber amplifier model based on a neural network purely trained with experimental data sets in E- and S-bands. This method allows a robust prediction of the amplifier operation that incorporates variations of fiber properties due to manufacturing process and any fluctuations of the amplifier characteristics. Using the proposed approach the spectral dependencies of gain and noise figure for given bi-directional pump currents and input signal powers have been obtained. The low mean (less than 0.19 dB) and standard deviation (less than 0.09 dB) of the maximum error are achieved for gain and noise figure predictions in the 1410–1490 nm spectral band.

A polarization-diversity loop with a silicon waveguide with a lateral p-i-n diode as a nonlinear medium is used to realize polarization insensitive four-wave mixing. Wavelength conversion of seven dual-polarization 16-quadrature amplitude modulation (QAM) signals at 16 GBd is demonstrated with an optical signal-to-noise ratio penalty below 0.7 dB. High-quality converted signals are generated thanks to the low polarization dependence (≤0.5 dB) and the high conversion efficiency (CE) achievable. The strong Kerr nonlinearity in silicon and the decrease of detrimental free-carrier absorption due to the reverse-biased p-i-n diode are key in ensuring the high CE levels.